nodemanager.c

ocfs1.2.7 源码

	*val = tmp;

	return count;
}

static ssize_t o2nm_cluster_attr_idle_timeout_ms_read(
	struct o2nm_cluster *cluster, char *page)
{
	return sprintf(page, "%u\n", cluster->cl_idle_timeout_ms);
}

static ssize_t o2nm_cluster_attr_idle_timeout_ms_write(
	struct o2nm_cluster *cluster, const char *page, size_t count)
{
	ssize_t ret;
	unsigned int val;

	ret =  o2nm_cluster_attr_write(page, count, &val);

	if (ret > 0) {
		if (cluster->cl_idle_timeout_ms != val
			&& o2net_num_connected_peers()) {
			mlog(ML_NOTICE,
			     "o2net: cannot change idle timeout after "
			     "the first peer has agreed to it."
			     "  %d connected peers\n",
			     o2net_num_connected_peers());
			ret = -EINVAL;
		} else if (val <= cluster->cl_keepalive_delay_ms) {
			mlog(ML_NOTICE, "o2net: idle timeout must be larger "
			     "than keepalive delay\n");
			ret = -EINVAL;
		} else {
			cluster->cl_idle_timeout_ms = val;
		}
	}

	return ret;
}

static ssize_t o2nm_cluster_attr_keepalive_delay_ms_read(
	struct o2nm_cluster *cluster, char *page)
{
	return sprintf(page, "%u\n", cluster->cl_keepalive_delay_ms);
}

static ssize_t o2nm_cluster_attr_keepalive_delay_ms_write(
	struct o2nm_cluster *cluster, const char *page, size_t count)
{
	ssize_t ret;
	unsigned int val;

	ret =  o2nm_cluster_attr_write(page, count, &val);

	if (ret > 0) {
		if (cluster->cl_keepalive_delay_ms != val
		    && o2net_num_connected_peers()) {
			mlog(ML_NOTICE,
			     "o2net: cannot change keepalive delay after"
			     " the first peer has agreed to it."
			     "  %d connected peers\n",
			     o2net_num_connected_peers());
			ret = -EINVAL;
		} else if (val >= cluster->cl_idle_timeout_ms) {
			mlog(ML_NOTICE, "o2net: keepalive delay must be "
			     "smaller than idle timeout\n");
			ret = -EINVAL;
		} else {
			cluster->cl_keepalive_delay_ms = val;
		}
	}

	return ret;
}

static ssize_t o2nm_cluster_attr_reconnect_delay_ms_read(
	struct o2nm_cluster *cluster, char *page)
{
	return sprintf(page, "%u\n", cluster->cl_reconnect_delay_ms);
}

static ssize_t o2nm_cluster_attr_reconnect_delay_ms_write(
	struct o2nm_cluster *cluster, const char *page, size_t count)
{
	return o2nm_cluster_attr_write(page, count,
	                               &cluster->cl_reconnect_delay_ms);
}
static struct o2nm_cluster_attribute o2nm_cluster_attr_idle_timeout_ms = {
	.attr	= { .ca_owner = THIS_MODULE,
		    .ca_name = "idle_timeout_ms",
		    .ca_mode = S_IRUGO | S_IWUSR },
	.show	= o2nm_cluster_attr_idle_timeout_ms_read,
	.store	= o2nm_cluster_attr_idle_timeout_ms_write,
};

static struct o2nm_cluster_attribute o2nm_cluster_attr_keepalive_delay_ms = {
	.attr	= { .ca_owner = THIS_MODULE,
		    .ca_name = "keepalive_delay_ms",
		    .ca_mode = S_IRUGO | S_IWUSR },
	.show	= o2nm_cluster_attr_keepalive_delay_ms_read,
	.store	= o2nm_cluster_attr_keepalive_delay_ms_write,
};

static struct o2nm_cluster_attribute o2nm_cluster_attr_reconnect_delay_ms = {
	.attr	= { .ca_owner = THIS_MODULE,
		    .ca_name = "reconnect_delay_ms",
		    .ca_mode = S_IRUGO | S_IWUSR },
	.show	= o2nm_cluster_attr_reconnect_delay_ms_read,
	.store	= o2nm_cluster_attr_reconnect_delay_ms_write,
};

static struct configfs_attribute *o2nm_cluster_attrs[] = {
	&o2nm_cluster_attr_idle_timeout_ms.attr,
	&o2nm_cluster_attr_keepalive_delay_ms.attr,
	&o2nm_cluster_attr_reconnect_delay_ms.attr,
	NULL,
};
static ssize_t o2nm_cluster_show(struct config_item *item,
                                 struct configfs_attribute *attr,
                                 char *page)
{
	struct o2nm_cluster *cluster = to_o2nm_cluster(item);
	struct o2nm_cluster_attribute *o2nm_cluster_attr =
		container_of(attr, struct o2nm_cluster_attribute, attr);
	ssize_t ret = 0;

	if (o2nm_cluster_attr->show)
		ret = o2nm_cluster_attr->show(cluster, page);
	return ret;
}

static ssize_t o2nm_cluster_store(struct config_item *item,
                                  struct configfs_attribute *attr,
                                  const char *page, size_t count)
{
	struct o2nm_cluster *cluster = to_o2nm_cluster(item);
	struct o2nm_cluster_attribute *o2nm_cluster_attr =
		container_of(attr, struct o2nm_cluster_attribute, attr);
	ssize_t ret;

	if (o2nm_cluster_attr->store == NULL) {
		ret = -EINVAL;
		goto out;
	}

	ret = o2nm_cluster_attr->store(cluster, page, count);
	if (ret < count)
		goto out;
out:
	return ret;
}

static struct config_item *o2nm_node_group_make_item(struct config_group *group,
						     const char *name)
{
	struct o2nm_node *node = NULL;
	struct config_item *ret = NULL;

	if (strlen(name) > O2NM_MAX_NAME_LEN)
		goto out; /* ENAMETOOLONG */

	node = kcalloc(1, sizeof(struct o2nm_node), GFP_KERNEL);
	if (node == NULL)
		goto out; /* ENOMEM */

	strcpy(node->nd_name, name); /* use item.ci_namebuf instead? */
	config_item_init_type_name(&node->nd_item, name, &o2nm_node_type);
	spin_lock_init(&node->nd_lock);

	ret = &node->nd_item;

out:
	if (ret == NULL)
		kfree(node);

	return ret;
}

static void o2nm_node_group_drop_item(struct config_group *group,
				      struct config_item *item)
{
	struct o2nm_node *node = to_o2nm_node(item);
	struct o2nm_cluster *cluster = to_o2nm_cluster(group->cg_item.ci_parent);

	o2net_disconnect_node(node);

	if (cluster->cl_has_local &&
	    (cluster->cl_local_node == node->nd_num)) {
		cluster->cl_has_local = 0;
		cluster->cl_local_node = O2NM_INVALID_NODE_NUM;
		o2net_stop_listening(node);
	}

	/* XXX call into net to stop this node from trading messages */

	write_lock(&cluster->cl_nodes_lock);

	/* XXX sloppy */
	if (node->nd_ipv4_address)
		rb_erase(&node->nd_ip_node, &cluster->cl_node_ip_tree);

	/* nd_num might be 0 if the node number hasn't been set.. */
	if (cluster->cl_nodes[node->nd_num] == node) {
		cluster->cl_nodes[node->nd_num] = NULL;
		clear_bit(node->nd_num, cluster->cl_nodes_bitmap);
	}
	write_unlock(&cluster->cl_nodes_lock);

	config_item_put(item);
}

static struct configfs_group_operations o2nm_node_group_group_ops = {
	.make_item	= o2nm_node_group_make_item,
	.drop_item	= o2nm_node_group_drop_item,
};

static struct config_item_type o2nm_node_group_type = {
	.ct_group_ops	= &o2nm_node_group_group_ops,
	.ct_owner	= THIS_MODULE,
};

/* cluster */

static void o2nm_cluster_release(struct config_item *item)
{
	struct o2nm_cluster *cluster = to_o2nm_cluster(item);

	kfree(cluster->cl_group.default_groups);
	kfree(cluster);
}

static struct configfs_item_operations o2nm_cluster_item_ops = {
	.release	= o2nm_cluster_release,
	.show_attribute		= o2nm_cluster_show,
	.store_attribute	= o2nm_cluster_store,
};

static struct config_item_type o2nm_cluster_type = {
	.ct_item_ops	= &o2nm_cluster_item_ops,
	.ct_attrs	= o2nm_cluster_attrs,
	.ct_owner	= THIS_MODULE,
};

/* cluster set */

struct o2nm_cluster_group {
	struct configfs_subsystem cs_subsys;
	/* some stuff? */
};

#if 0
static struct o2nm_cluster_group *to_o2nm_cluster_group(struct config_group *group)
{
	return group ?
		container_of(to_configfs_subsystem(group), struct o2nm_cluster_group, cs_subsys)
	       : NULL;
}
#endif

static struct config_group *o2nm_cluster_group_make_group(struct config_group *group,
							  const char *name)
{
	struct o2nm_cluster *cluster = NULL;
	struct o2nm_node_group *ns = NULL;
	struct config_group *o2hb_group = NULL, *ret = NULL;
	void *defs = NULL;

	/* this runs under the parent dir's i_sem; there can be only
	 * one caller in here at a time */
	if (o2nm_single_cluster)
		goto out; /* ENOSPC */

	cluster = kcalloc(1, sizeof(struct o2nm_cluster), GFP_KERNEL);
	ns = kcalloc(1, sizeof(struct o2nm_node_group), GFP_KERNEL);
	defs = kcalloc(3, sizeof(struct config_group *), GFP_KERNEL);
	o2hb_group = o2hb_alloc_hb_set();
	if (cluster == NULL || ns == NULL || o2hb_group == NULL || defs == NULL)
		goto out;

	config_group_init_type_name(&cluster->cl_group, name,
				    &o2nm_cluster_type);
	config_group_init_type_name(&ns->ns_group, "node",
				    &o2nm_node_group_type);

	cluster->cl_group.default_groups = defs;
	cluster->cl_group.default_groups[0] = &ns->ns_group;
	cluster->cl_group.default_groups[1] = o2hb_group;
	cluster->cl_group.default_groups[2] = NULL;
	rwlock_init(&cluster->cl_nodes_lock);
	cluster->cl_node_ip_tree = RB_ROOT;
	cluster->cl_reconnect_delay_ms = O2NET_RECONNECT_DELAY_MS_DEFAULT;
	cluster->cl_idle_timeout_ms    = O2NET_IDLE_TIMEOUT_MS_DEFAULT;
	cluster->cl_keepalive_delay_ms = O2NET_KEEPALIVE_DELAY_MS_DEFAULT;

	ret = &cluster->cl_group;
	o2nm_single_cluster = cluster;

out:
	if (ret == NULL) {
		kfree(cluster);
		kfree(ns);
		o2hb_free_hb_set(o2hb_group);
		kfree(defs);
	}

	return ret;
}

static void o2nm_cluster_group_drop_item(struct config_group *group, struct config_item *item)
{
	struct o2nm_cluster *cluster = to_o2nm_cluster(item);
	int i;
	struct config_item *killme;

	BUG_ON(o2nm_single_cluster != cluster);
	o2nm_single_cluster = NULL;

	for (i = 0; cluster->cl_group.default_groups[i]; i++) {
		killme = &cluster->cl_group.default_groups[i]->cg_item;
		cluster->cl_group.default_groups[i] = NULL;
		config_item_put(killme);
	}

	config_item_put(item);
}

static struct configfs_group_operations o2nm_cluster_group_group_ops = {
	.make_group	= o2nm_cluster_group_make_group,
	.drop_item	= o2nm_cluster_group_drop_item,
};

static struct config_item_type o2nm_cluster_group_type = {
	.ct_group_ops	= &o2nm_cluster_group_group_ops,
	.ct_owner	= THIS_MODULE,
};

static struct o2nm_cluster_group o2nm_cluster_group = {
	.cs_subsys = {
		.su_group = {
			.cg_item = {
				.ci_namebuf = "cluster",
				.ci_type = &o2nm_cluster_group_type,
			},
		},
	},
};

#define O2NM_PROC_PATH "fs/ocfs2_nodemanager"
static struct proc_dir_entry *o2nm_proc;

#define O2NM_VERSION_PROC_NAME "interface_revision"
#define O2NM_HB_DEAD_THRESHOLD_NAME "hb_dead_threshold"
#define O2NM_FENCE_OPTION_NAME "fence_method"

static void o2nm_remove_proc(struct proc_dir_entry *parent)
{
	remove_proc_entry(O2NM_VERSION_PROC_NAME, parent);
	remove_proc_entry(O2NM_HB_DEAD_THRESHOLD_NAME, parent);
	remove_proc_entry(O2NM_FENCE_OPTION_NAME, parent);
}

static void __exit exit_o2nm(void)
{
	if (ocfs2_table_header)
		unregister_sysctl_table(ocfs2_table_header);

	/* XXX sync with hb callbacks and shut down hb? */
	o2net_unregister_hb_callbacks();
	configfs_unregister_subsystem(&o2nm_cluster_group.cs_subsys);
	o2nm_remove_proc(o2nm_proc);
	mlog_remove_proc(o2nm_proc);
	o2net_proc_exit(o2nm_proc);
	remove_proc_entry(O2NM_PROC_PATH, NULL);

	o2net_exit();
}

static int o2nm_proc_write_uint(struct file *file, const char __user *buffer,
				unsigned long count, void *data)
{
	char buf[32];
	char *p = buf;
	unsigned long tmp;

	if (count > ARRAY_SIZE(buf) - 1)
		count = ARRAY_SIZE(buf) - 1;

	if (copy_from_user(buf, buffer, count))
		return -EFAULT;

	buf[ARRAY_SIZE(buf) - 1] = '\0';

	tmp = simple_strtoul(p, &p, 10);
	if (!p || (*p && (*p != '\n')))
                return -EINVAL;

	*(unsigned int *)data = tmp;

	return count;
}

static int o2nm_proc_read_uint(char *page, char **start, off_t off,
			       int count, int *eof, void *data)
{
	int len;
	unsigned int val = *(unsigned int *)data;

	len = sprintf(page, "%u\n", val);
	if (len < 0)
		return len;

	if (len <= off + count)
		*eof = 1;

	*start = page + off;

	len -= off;

	if (len > count)
		len = count;

	if (len < 0)
		len = 0;

	return len;
}

static int o2nm_proc_write_threshold(struct file *file,
				     const char __user *buffer,
				     unsigned long count, void *data)
{
	char buf[32];
	char *p = buf;
	unsigned long tmp;

	if (count > ARRAY_SIZE(buf) - 1)
		count = ARRAY_SIZE(buf) - 1;

	if (copy_from_user(buf, buffer, count))
		return -EFAULT;

	buf[ARRAY_SIZE(buf) - 1] = '\0';

	tmp = simple_strtoul(p, &p, 10);
	if (!p || (*p && (*p != '\n')))
                return -EINVAL;

	/* this will validate ranges for us. */
	o2hb_dead_threshold_set((unsigned int) tmp);

	return count;
}

static int o2nm_init_proc(struct proc_dir_entry *parent)
{
	struct proc_dir_entry *p;

	p = create_proc_entry(O2NM_VERSION_PROC_NAME,
			      S_IFREG | S_IRUGO, parent);
	if (!p)
		goto bail1;
	p->read_proc = o2nm_proc_read_uint;
	p->data = (void *)&o2nm_api_version;

	p = create_proc_entry(O2NM_HB_DEAD_THRESHOLD_NAME,
			      S_IFREG | S_IRUGO | S_IWUSR, parent);
	if (!p)
		goto bail2;
	p->write_proc = o2nm_proc_write_threshold;
	p->read_proc = o2nm_proc_read_uint;
	p->data = (void *)&o2hb_dead_threshold;

	p = create_proc_entry(O2NM_FENCE_OPTION_NAME,
			      S_IFREG | S_IRUGO | S_IWUSR, parent);
	if (!p)
		goto bail3;
	p->write_proc = o2nm_proc_write_uint;
	p->read_proc = o2nm_proc_read_uint;
	p->data = (void *)&o2quo_fence_option;

	return 0;

bail3:
	remove_proc_entry(O2NM_HB_DEAD_THRESHOLD_NAME, parent);
bail2:
	remove_proc_entry(O2NM_VERSION_PROC_NAME, parent);
bail1:
	return -ENOMEM;
}

static int __init init_o2nm(void)
{
	int ret = -1;

	cluster_print_version();

	o2hb_init();
	o2net_init();

	ocfs2_table_header = register_sysctl_table(ocfs2_root_table, 0);
	if (!ocfs2_table_header) {
		printk(KERN_ERR "nodemanager: unable to register sysctl\n");
		ret = -ENOMEM; /* or something. */
		goto out;
	}

	ret = o2net_register_hb_callbacks();
	if (ret)
		goto out_sysctl;

	config_group_init(&o2nm_cluster_group.cs_subsys.su_group);
	init_MUTEX(&o2nm_cluster_group.cs_subsys.su_sem);
	ret = configfs_register_subsystem(&o2nm_cluster_group.cs_subsys);
	if (ret) {
		printk(KERN_ERR "nodemanager: Registration returned %d\n", ret);
		goto out_callbacks;
	}

	o2nm_proc = proc_mkdir(O2NM_PROC_PATH, NULL);
	if (o2nm_proc == NULL) {
		ret = -ENOMEM; /* shrug */
		goto out_subsys;
	}

	ret = mlog_init_proc(o2nm_proc);
	if (ret)
		goto out_remove;

	ret = o2net_proc_init(o2nm_proc);
	if (ret)
		goto out_mlog;

	ret = o2nm_init_proc(o2nm_proc);
	if (ret == 0)
		goto out;

out_mlog:
	mlog_remove_proc(o2nm_proc);
out_remove:
	remove_proc_entry(O2NM_PROC_PATH, NULL);
out_subsys:
	configfs_unregister_subsystem(&o2nm_cluster_group.cs_subsys);
out_callbacks:
	o2net_unregister_hb_callbacks();
out_sysctl:
	unregister_sysctl_table(ocfs2_table_header);
out:
	return ret;
}

MODULE_AUTHOR("Oracle");
MODULE_LICENSE("GPL");

module_init(init_o2nm)
module_exit(exit_o2nm)